The adverse impact of early-onset conduct problems on both the individuals engaging in conduct problems and society makes the examination of the etiology of conduct problems extremely important. The "developmental propensity model" (Lahey &Waldman, 2003) advances specific and testable hypotheses regarding the etiology of conduct problems, proposing three dimensions of temperament (i.e., negative emotionality, daring, and low prosociality) and low cognitive ability as distinct components of "antisocial propensity". The proposed study will test the developmental propensity model and alternative hypotheses Keenan &Shaw, 2003;Nigg &Huang-Pollock, 2003), examining very early predictors of early-onset conduct problems and related disruptive behavior disorders. Specifically, the proposed study will: 1) Examine the bur components of "antisocial propensity" as early predictors of early-onset conduct problems and related disruptive behavior disorders, 2) Test specific hypotheses regarding sex differences in conduct problems, 3) Examine the genetic influences on antisocial propensity and conduct problems, 4) Examine the influence of negative parenting/family environment on conduct problems, and 5) Test specific hypotheses regarding comorbidity in conduct problems. To achieve the above aims, the proposed study will examine existing data from the Colorado Longitudinal Twin Study (LTS), a unique resource of 410 pairs of monozygotic and dizygotic same-sex twins who have been assessed since age 14 months. The LTS affords an exceptional opportunity to test the developmental propensity model, given its longitudinal, genetically informative data. Data analyses being proposed will focus on: a) parenting ratings, teacher ratings, and observational measures of antisocial propensity collected from age 14 months to 7 years, b) parent and teacher ratings of conduct problems from age 4 to 12, c) self ratings of conduct problems collected at age 12, and d) self ratings of conduct problems collected at age 16-19. A substantial amount of work remains in order to make full use of the great potential of the LTS dataset for testing the above specific aims. The proposed study will a) transfer approximately 20,000 hours of observational data from videotapes to DVDs, b) complete the coding of more than 10,000 observational tests of antisocial propensity and parenting, c) systematically review all existing coding for completeness and accuracy of data entry, and d) integrate the newly coded data and the existing data. The proposed study will provide the first comprehensive test of disconfirmable hypotheses from the developmental propensity model for the etiology of antisocial behavior in a genetically informative, multivariate, longitudinal study. As such, the proposed research will permit critical tests of hypotheses derived from the model, and thereby provide important new knowledge regarding the etiology of antisocial behavior, an issue of considerable societal and clinical significance.